Sending identical messages to those who replied, those who read and stayed silent, and those who never received it - common when rushing. Result is predictable: happy clients get annoyed by repeats, silent ones report spam. How WhatsApp technical statuses become real segments and why Excel is a logging tool, not a data source.
WhatsApp returns four technical message statuses: sent (reached server), delivered (reached device), read (opened by recipient), and failed (delivery error). Real system events underpinning automatic database tagging in CRM and mailing services - if the base was built correctly from the start, as in incoming leads breakdown.
Each status needs a different next step, not the same message on repeat.
Before:
Different messages per segment boost response 2–3x.
After:
Per agency cases, customized offer for "read but no reply" segment - objection handling or bonus - raises conversion 2–3x vs resending same text to entire base. No official Meta confirmation - practical benchmark, not documented standard.
Before:
How to tag the base directly in Excel.
After:
Excel is a static table - it doesn't receive delivery/read statuses on its own. Dynamic segmentation needs CRM or WABA API catching Meta webhooks; Excel gets exported logs for filtering and analytics.
Additional original inaccuracy:
"Delivered" = "didn't read."
Wrong. If recipient disabled read receipts in privacy settings, read never forms even after real reading - message stays delivered forever. Part of active audience lands in "didn't read" segment due to privacy settings, not real ignoring.
| Status | Technical meaning | Next step |
|---|---|---|
| Read, no reply | Opened, no inbound from JID within timeout | Handle objection, bonus, clarifying question - not same text repeat |
| Delivered, no read | Reached device, read status not formed (including disabled receipts) | Duplicate channel (SMS) or soft reminder, no aggression |
| Failed | Usually number not registered on WhatsApp | Remove from base - resend wastes limits |
| Replied | Two-way dialog exists | Exclude from bulk blast - continue only in dialog context |
Not three segments from original thesis - minimum four; failed deserves separate handling.
Dynamic segmentation starts not in Excel but in a system catching statuses:
In all three cases Excel stores and filters received statuses - not the collection source.
E-commerce marketer exported first wave report to CSV, VLOOKUP-filtered in Excel: failed removed as invalid, read no reply tagged "Segment B" with 10% discount reminder, delivered no read - "Segment C" with SMS duplicate offer. Repeat-touch revenue up 140%, no blocks recorded.
Shows: splitting base on at least two axes (read/didn't read) with different messages per axis gives measurable gain without harming number. Doesn't prove: 140% reproducible in any niche - one documented result.
Opposite: beauty salon sent promo to 3000-number base, same text again 3 days later to entire base including those who booked or clearly ignored. Repeat perceived as spam - mass complaints, gray work number banned at 200 sends in wave two. Difference - segmentation before repeat touch, not technology - see mass mailing ban mechanics.
Practitioners disagree on timing. Some say follow-up 2–4 hours after read while attention still on chat. Others argue that speed looks like bot stalking and triggers complaints - safe interval strictly 24–48 hours in their view.
No confirmation either way - practical preference, not documented rule - like behavioral antispam guides.
Confirmed activity segmentation guides: active - opened in last month; inactive - no activity 3+ months; "last read within 60 days" - practice for contacts who opened chat in last 2 months. Excel integration via ready services - ~5 minutes setup with up to 30-day trial.
Forum, undocumented Meta metrics: 2–3x conversion lift personalizing "read no reply" vs same text to whole base; 15–25% users disable blue read ticks, systematic error in manual segment export; behavioral slices reduce complaints 40–60% per practitioner observations.
"Delivered means definitely didn't read"
Wrong. Read via notification shade, preview mode, or disabled read receipts - all stay delivered.
"Excel tracks reads in real time"
Wrong. Static table; dynamic read/delivered segmentation needs webhook-catching system; Excel gets final log only.
"No reply = send same text again" Wrong. Repeat without context change = spam - beauty salon case proves it.
"Failed = try resending tomorrow" Wrong. Usually number not on WhatsApp - resend wastes limits.
"Read no reply = not interested" Wrong as universal rule. Busy, deferred decision - objection-handling segment, not write-off.
failed segment needs cleaning, not re-attacking - usually dead number.If mailings go one text to whole base now - start simple: export last campaign statuses, split at least read/didn't read. Already reduces re-irritating those who saw message and didn't reply.
Practical rule:
One text for the whole base is a message not to the client but to yourself: "I can't be bothered to see who reacted how."